The Imperative For Unbanning The USB
USB devices have caused some of the most preventable data loss events in the last decade. In June it transpired that counterfeit USB drives carrying malware thought to be linked to Chinese state actors had been used to infect Japanese military networks for almost a year.
Two years ago, we also saw the GoldenJackal group develop malware configured specifically for USB drive so that any drive infected was then able to capture data when connected to an air-gapped system and relay it to the group when it was once again inserted into an internet-connected computer.
And we’ve all heard tales of USB drops whereby bad USBs were used to dupe individuals into using them, whether dropped in car parks or mailed directly to recipients working for targeted industries.
Yet there have also been notable instances where USB sticks have saved the day. When the Industrial & Commercial Bank of China’s US division was hit by a cyber attack in 2023, the bank was able to continue trading by sending settlement details for transactions on USB keys via courier.
USB devices have also been used to enable displaced or nomadic populations to carry their health records with them, an example being Sijilli, a Lebanese cloud-based system devised to assist the Syrian refugee patient population. Plus, they are of course used every day to transport data over air-gapped systems, in OT/ICS environments, over forensic workflows, in field operations or on cross-domain networks.
To Ban or Not to Ban?
It’s a dichotomous situation then and one that has seen some organisations resort to blanket bans. However, it has since become apparent that while blocking portable media stops the threat, it also hamstrings the business. Consequently, many CISOs now champion the need for an ‘allow list’ of enterprise-grade devices provided they pass muster in that they are secure, compliant and auditable. Because strictly speaking it’s not the concept of portable media that is the problem, but how they are built.
Unencrypted devices can easily be compromised if they become lost or stolen or can be used as a workaround to bypass policy constraints if insufficient controls are put in place.
When considering the criteria for that ‘allow list’ it’s therefore crucial to look at how USBs have been abused in the past and to factor in controls to mitigate that risk. Hardware-encrypted devices, for example, are preferable in that they feature on-device authentication, such as via a keypad for PIN entry, and make encryption mandatory at the silicon layer. They also frequently have tamper resistant key storage that prevents brute forcing or other hacking or reverse engineering techniques. And relying on hardware as opposed to software encryption with on-device authentication also enables the device to be used with endpoints that have been locked down because there’s no need to download drivers.
The firmware for the device also needs to be “secure” – as in it is locked down and cannot be modified in any way to be used to launch an attack such as those previously highlighted.
USBs & Zero Trust
Similarly, the USB device needs to be able to fit in with a modern zero trust architecture. A common misconception around Zero Trust is that its default position is to deny all. That’s not strictly true as ZTA simply requires verification and so the position is actually only trust a device if it can prove that it is trustworthy. To do that requires a verifiable identity, enforced encryption and auditable usage.
Another key consideration is compliance. GRC teams prefer devices that adhere to known standards and controls. With respect to USBs, that equates to FIPS 140 2/3 validation which ensures the device has tamper-proof, identity-based cryptographic encryption that uses approved algorithms. The standard is deliberately rigorous because it applies to devices aimed at non-military US government agencies and contractors so sets the bar high, making the standard also desirable to security-conscious enterprises. Deterministic encryption and hardware based key management are also advantageous as these provide auditors with provable, certifiable controls.
Cost Versus consequence
Will there be a higher cost associated with these devices that make the ‘allow list’? Yes, when compared to a plug-and-play USB stick. But while a secure device may cost £50-100 more than a commodity drive, any security breach is likely to dwarf that by a factor of a thousand - and that’s not counting the reputational damage. When it comes to weighing up the cost versus risk, there’s really no contest.
It’s important to note, however, that what CISOs aren’t doing is arguing for the return of the USB per se. They’re not looking to turn back the clock or resurrect an old technology. Instead, they’re looking to provide their workforce with data mobility and the means to get the job done.
And to do that they need secure, controlled and policy-enforced devices which actually differ significantly from those commodity USBs. In fact, those that do make the grade effectively eliminate the reasons for why USBs were banned in the first place.
Jon Fielding is Managing Director, EMEA at Apricorn
Image: GetLostMike
You Might Also Read:
Are Organisations Finally Grasping The Importance Of Encryption?:
If you like this website and use the comprehensive 8,000-plus service supplier Directory, you can get unrestricted access, including the exclusive in-depth Directors Report series, by signing up for a Premium Subscription.
- Individual £5 per month or £50 per year. Sign Up
- Multi-User, Corporate & Library Accounts Available on Request
- Inquiries: Contact Cyber Security Intelligence
Cyber Security Intelligence: Captured Organised & Accessible






